16:01:01 <noonedeadpunk> #startmeeting openstack_ansible_meeting
16:01:02 <openstack> Meeting started Tue Dec 10 16:01:01 2019 UTC and is due to finish in 60 minutes.  The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:06 <openstack> The meeting name has been set to 'openstack_ansible_meeting'
16:01:13 <guilhermesp> o/
16:01:17 <noonedeadpunk> #topic office hours
16:01:27 <gshippey> o/
16:01:42 <raukadah> \o/
16:02:37 <noonedeadpunk> so from my side I think we're ready for releasing and actually we _must_ release
16:03:14 <noonedeadpunk> just few things left to merge and I've already placed a bump on top of them
16:03:40 <noonedeadpunk> IF anyone has any objections or not addressed issues it's better to say them loud now :)
16:03:56 <guilhermesp> hum so we need that 4 patches merged then
16:06:48 <noonedeadpunk> actualy 5 (1 is cherry-picked while master is not merged
16:07:13 <noonedeadpunk> IF you think we should drop off some out of the process (not to wait for it) - we can do that
16:11:41 <guilhermesp> i think we can drop off for now https://review.opendev.org/#/c/698262/ if is not crucial
16:13:54 <noonedeadpunk> you mean that we can offer ppl to do upgrades without using upgrade script?
16:14:06 <noonedeadpunk> Like if they use non-standart directories?
16:14:13 <mnaser> i think guilhermesp means that the OSA_CONFIG_DIR stuff is an extra
16:14:27 <mnaser> i know we use it but the change is trivial so probably doesnt hurt that much, unless its really a pita to merge then we can drop it
16:14:29 <guilhermesp> yep
16:15:20 <mnaser> epel timing out has hurt centos i guess
16:15:25 <noonedeadpunk> Yeah
16:15:34 <mnaser> does infra have an epel mirror?
16:15:40 <noonedeadpunk> But this happened to several patches
16:15:54 <noonedeadpunk> I think they did...
16:16:30 <mnaser> http://mirror.sjc1.vexxhost.openstack.org/epel/
16:16:31 <mnaser> they do
16:16:33 <mnaser> i think we use it?
16:16:47 <mnaser> i think the problem here is the fact the GPG key is being downloaded from the internes
16:16:51 <mnaser> intarwebs
16:17:22 <mnaser> we should install this key locally
16:17:26 <mnaser> i can submit a patch for this..
16:19:20 <noonedeadpunk> that would be nice I think
16:20:22 <noonedeadpunk> Ok, one more thing I'd love to hear opinion about is https://review.opendev.org/#/c/697830/ I dropped it from osa-train topic as this is pretty crucial patch
16:21:31 <noonedeadpunk> And I didn't have a chance to do some testing (with several control nodes) against master and train, but on rocky and stein one failed memcached still brings problems and latency
16:26:04 <openstackgerrit> Mohammed Naser proposed openstack/openstack-ansible-repo_server master: epel: use vendored gpg keys  https://review.opendev.org/698290
16:26:31 <mnaser> noonedeadpunk: yeah, i think this is somethign we need more testing on, its a pretty fundamental change, some OSA users use an external LB too
16:28:46 <noonedeadpunk> actually it's pretty overridable now, so this can be moved to some docs instead of changing default behaviour
16:32:09 <mnaser> id be ok with that
16:35:46 <noonedeadpunk> I think we should do https://review.opendev.org/698290 in some other way, as there's a folk who's using some another repo and was using centos_epel_key and  centos_epel_mirror to override stuff across all roles
16:51:06 <mnaser> noonedeadpunk: wanna leave a review there?
16:52:45 <noonedeadpunk> yeah, will do:)
16:53:01 <noonedeadpunk> just what to think about what better solution to offer instead:)
16:54:26 <cjloader> o/
16:54:32 <cjloader> sorry late
17:00:14 <raukadah> noonedeadpunk: jrosser mnaser https://review.opendev.org/#/c/698134/ packstack review might be useful for OSA
17:01:01 <mnaser> raukadah: yay el8 pkgs?
17:01:14 <raukadah> mnaser: yes
17:01:29 <noonedeadpunk> cool!
17:01:30 <raukadah> mnaser: not complete yet, but half baked might be useful
17:01:38 <noonedeadpunk> #endmeeting